Laser probe microanalysis: performance assessment and applications

Abstract

The LIMA 2A instrument used for this work has been described in detail elsewhere.' A focused frequency-quadrupled Nd-YAG laser incident on the surface of a solid sample, via Cassegrain optics, provides a burst of ions for analysis in a time-of-flight mass spectrometer. This consists of a folded field-free flight tube with an electrostatic reflector; designed to compensate for the initial spread in ion energies,and a modified venetian-blind electron multiplier for ion detection. The current transient at the output of the electron multiplier is buffered and stored in a fast transient recorder for further processing. Material is sampled from a crater 2-3 p m in diameter and -0.1 p n i in depth with the option of a crude depth profile provided by repeated shots at the same position. As yet, in common with other laser microprobes,3 the resulting mass-peak areas give only semiquantitative information on the relative abundances of elements in a sample. For many problems this information is sufficient and LIMA has proved a valuable tool in thc investigation of the surface and near-surface conditions in a variety of samples. due to the rapid rate at which data is obtained together with the minimal specimen preparation required; some specific examples are detailed later. Efforts have also been made to identify factors responsible for the variable elemental sensitivities and the scatter in cxperimental results. OUANTIFIC ATIONThe measured peak area of a particular element depends primarily on its concentration, the ionization potential, the absorbed laser eneregy and a machine transmission/detection efficiency? To obtain quantitative analyses these dependences must first be estimated either from theoretical models or empirically, but in either case, consistent experimental results are required. The dependence on absorbed laser energy is strong and although the incident laser energy at a constant setting has a shot-to-shot variation of only 3% (one standard deviation, 1 C J ) , the measured scatter in absolute peak areas for Si ranges from about 130% (1 (7) at low laserpower to about 10% at about double this power. Most of this scatter is probably due to variations in the efficiency with which the laser energy is absorbed. The scatter is reduced if peak-area ratios are used, but differences in the dependence of ion yield on absorbed power from element to element and the presence of non-lincarities in the detection system mean that the scatter is still large, especially at low laser-power.One approach to quantification is to use a high laser-power, where the relative sensitivities appear to saturate. and an empirical relationship for sensitivity vs ionization potential.4 However, the penalty is a loss in spatial rewlution with craters as big as 20 pin in diameter. If two different elemental ratios, measured at various laser powers on a homogeneous specimen, are plotted against each other a curve is obtained which suggests that the measured ratio of two elements of known concentration could be used to predict a correction fact...